Analysis Movement of the Sailboat Mondisy Figure1. Mondisy track (not to scale) From the nearby cliffs and looking down at the Mondisy, the sailboat's true movement could be observed, that is circling. However, from a distance such as on board the CanadaSenator, the circular movements of the Mondisy appeared as a zigzag manoeuvre. Repeatedly turning in circles is consistent with either a mechanical malfunction related to the steering, or the helm simply being left unattended for some time before the collision. There was no sign of activity on deck during the 15-minute period before the collision. Ferro-cement Construction Although reports of pleasure craft hindering large vessels in the main shipping channel is a common occurrence, collisions are rare. In1986, a 12m fibreglass sailboat and a laker were involved in a collision in the main shipping channel near Chippewa Point, New York, United States.8 The sailboat remained perched on the bulbous bow of the laker for several minutes and, when way was taken off the laker, the sailboat slipped off and remained afloat. The sailboat sustained only minor hull damage and the occupants sustained some bruises. The Mondisy, of ferro-cement construction, was made of materials without inherent buoyancy and relatively unresilient. Upon impact with the steel bow of the CanadaSenator, the ferro-cement hull was breached and the sailboat quickly sank. Watchkeeping Schedules and Fatigue The crew of the sailboat relied on an unstructured single-person watchkeeping system that required the watchkeeper to also act as the helmsman/lookout and did not provide time for scheduled rest. Individuals were temporarily relieved from duty when they requested a break or when the owner of the sailboat felt he should operate the boat through difficult areas. The unstructured watchkeeping continued throughout the evening and night. At about 0400, the crew member requested relief from the co-worker. The fact that the crew member fell asleep in the companionway moments after being relieved is an indication that he had reached a level of fatigue that made him unfit for duty. At the time of the occurrence, the lookout and helm duties had been conferred on the co-worker. By this time, the co-worker had accumulated some 18hours of on-duty, off-duty activity on board the sailboat as well as some four to six hours of morning preparation time before departure; this translates into being awake continuously for at least 22hours. After long periods of continuous wakefulness, sleepiness becomes profound in the early morning hours.9 When coupled with a monotonous task, such as steering a boat essentially alone (all others were sleeping) and the resounding rhythm of the motor, falling asleep becomes difficult to avoid. There is no information to suggest that alcohol was consumed during the trip and because the co-worker's toxicology result was negative, the impact of drugs on his performance can be ruled out. It is therefore likely that the unstructured watchkeeping system aboard the Mondisy led to a sleep debt that caused the co-worker to fall asleep while on duty at the helm. Once asleep, he would have lost muscle tone and hence released the steering wheel. The vessel would then eventually have begun to track in a closed circle in the direction of the last rudder input. In this case, the wheel was turned counterclockwise, making the vessel turn to port. A recent study has indicated that watchkeeping schedules employing two or less people can lead to accidents caused by fatigue and that poor watchkeeping standards on commercial vessels decreases the likelihood of detecting collision risks.10 The risks due to improper watchkeeping practices on pleasure craft may be even greater than those found on commercial vessels because pleasure craft operators are neither required to follow any prescribed watchkeeping standard nor have they necessarily acquired proper training or awareness of these issues. Ship-to-Ship Communications Ship-to-ship communications are primarily carried out on very high frequency (VHF) radiotelephone frequencies. Notwithstanding this type of communication, vessels are also required to communicate using light and sound signalling devices pursuant to the International Regulations for Preventing Collisions at Sea (Collision Regulations). Radio communications between ships and with local MCTS centres are carried out by continuous monitoring of dedicated radiotelephone frequencies. When a vessel's intentions are found to be ambiguous, a Security radio message can be transmitted or relayed through a MCTS centre directly to the bridge team members concerned, thereby alerting the vessel to impending danger. Commercial vessels do not readily make use of VHF radiotelephones to communicate with pleasure craft in spite of the availability of these instruments on board many pleasure craft such as the Mondisy. As radio watchkeeping is not mandatory aboard pleasure craft, operators of commercial vessels are not inclined to communicate with pleasure craft by this means, as in this instance. More often than not, sound signals are used. Light and sound signals have limitations that decrease their effectiveness. For instance, light signals are less effective during the day. Local engine noise and/or an enclosed cockpit/bridge makes hearing of sound signals difficult, thereby reducing their effectiveness. In this instance, the sailboat's engine was operated throughout the voyage down river. The engine noise, together with the enclosed cockpit of the Mondisy, would have made the two warning signals by the CanadaSenator difficult to hear. Operational Environment and Safety Experience has shown that pleasure craft keep out of the way of large vessels that can only navigate within the confines of the channel. This expectation is further reinforced by the Collision Regulations11 that require them to do so. For many pilots, the sounding of the warning signal, that is at least five short and rapid blasts on the ship's whistle, usually produces the desired effect of getting attention and appropriate action. Commercial vessels often meet pleasure craft that are being operated in an apparently erratic and unpredictable manner in the main navigation channel. Under such conditions, it is often difficult to assess whether a risk of collision exists12 and substantial course alterations or speed reductions are seldom necessary or even possible in the time available. Action by Stand-on Vessel Under the Collision Regulations, the CanadaSenator was required to maintain its course and speed, but as soon as it became apparent that the Mondisy was not taking appropriate action to keep out of the way, the CanadaSenator was required to take such action as would be best to avoid collision.13 In this case, course alterations were carried out, but the CanadaSenator was not slowed down either to give it more time to assess the situation14 or when collision was imminent. Technological Advances The sailboat was equipped with an ECS programmed on a laptop computer and linked to a GPS navigation device. The navigational channel below Trois-Rivires is relatively difficult and the use of an ECS or effective use of radar is valuable for navigating safely through the channel. The crew lacked knowledge of night navigation using aids such as a radar. However, the presence of an ECS provided the crew a means to navigate within the confines of a channel, especially at night due to its relative ease of monitoring the vessel's position and in accordance with the operator's directions. The decision to undertake the long passage that included night navigation ought to have taken into consideration, among others, the ability to effectively monitor the vessel's progress, density and movement of traffic in the vicinity, and the fact that sufficient well-rested personnel are on watch (effective watchkeeping system). In this occurrence, the owner lost his life, and this precluded the investigation from ascertaining the factors that influenced his decision making to undertake this trip without stopping over for the night. Additionally, an operator lacking experience and training in navigation is likely to be overreliant on the ECS, a potentially risky situation when operating in restricted waterways at night or in poor visibility. Pleasure Craft Operator Training Of the three persons operating the Mondisy during the trip down river, only the owner held a valid PCOC. The crew member, being 18years of age, was required to, but did not have this proficiency. Keeping a proper lookout and ensuring that a well-rested crew operate the vessel makes good common sense; the principle of rest and maintaining a vigilant lookout for oncoming traffic is intuitive. Nonetheless, research has shown that fatigued individuals do not reliably estimate their own level of alertness and performance.15 Effective training, such as afforded through the PCOC, can help disseminate best practices and reinforce good seamanship practices. However, PCOC certification is not mandatory for the majority of pleasure craft operators until2009. Proper voyage planning, including the planning of rest periods for the crew to operate the vessel safely at all times, is not presently part of the PCOC syllabus. The crew of the Mondisy did not fully appreciate the benefits of keeping a good lookout, both auditive and visual,16 as well as the risk of fatigue when under way for an extended period of time without a structured relief system. Post-Collision Safety Risks Appropriate decision making and use of proper rescue equipment are critical to avoid additional risks after an accident occurs. Although not causal factors in this occurrence, the following situations had the possibility of exacerbating an already poor situation: After the collision, the survivors found themselves in the water without personal flotation devices. The liferaft, uninflated and floating on the surface of the water, was approached by one of the survivors but he was unable to deploy it because the painter proved to be too long to work with while swimming. The deployment painter must be pulled to activate the compressed air bottle. Safety issues with respect to liferaft inflation have been addressed in past TSB reports.17 During abandonment, the cumulative effect of anxiety resulting from the situation and the loss of valuable time associated with hauling the length of painter to inflate the liferaft may decrease the survival time of a non-swimmer. The CanadaSenator's man overboard lifebuoys and smoke flares were not released and the rescue boat was not operational within an effective time period.18 Although the collision took place at approximately 0554, the CanadaSenator's rescue boat was not lowered into the water until 0654, one full hour after the occurrence. Such a delay has the potential to reduce the chances of survival of persons in the water. A CCG SAR vessel was in the area by 0627. The CanadaSenator's rescue boat should have been launched once the vessel slowed down to bare steerageway. The pilot was not immediately replaced by a substitute once the CanadaSenator was safely anchored near the collision site. In the absence of a formal post-accident relief policy by the Laurentian Pilotage Authority and/or the Corporation des pilotes du Saint-Laurent central, the pilot chose to remain on the CanadaSenator during the SAR operation and then to weigh anchor, navigate down river, and then re-anchor the vessel at the port in Qubec; a process that took over three hours to complete. This is not an unreasonable amount of time under normal circumstances; however, in this instance, the pilot reported being exhausted by the time he disembarked. A relief was not requested. In previous reports and in a safety communication addressed to the Laurentian Pilotage Authority in 2001, the TSB emphasized that the lack of a formal post-accident pilot relief policy increases risks.19 This risk has been recognized by the Great Lakes Pilotage Authority which has had such a policy for some time. Collision with another vessel whose occupants are thrown into the water and with two persons missing is a very stressful event. Functioning under stress can cause decrements in performance such as narrowing of attention, reduction in working memory, and decrease in cognitive processing speed.20 These decrements are known precursors to increased risk of human error. It is likely that the unstructured watchkeeping system aboard the Mondisy led to a sleep debt that caused the co-worker to fall asleep while on duty at the helm. The warning signals sounded on the ship's whistle of the CanadaSenator did not produce the desired effect, that is arousing attention and indicating to the pleasure craft that it was standing into danger. Alterations of course without reduction in speed did not prevent the collision. The Mondisy, of ferro-cement construction, was of materials without inherent buoyancy and relatively unresilient. Upon impact with the CanadaSenator, the hull was breached and the sailboat quickly sank.Findings as to Causes and Contributing Factors It is likely that the unstructured watchkeeping system aboard the Mondisy led to a sleep debt that caused the co-worker to fall asleep while on duty at the helm. The warning signals sounded on the ship's whistle of the CanadaSenator did not produce the desired effect, that is arousing attention and indicating to the pleasure craft that it was standing into danger. Alterations of course without reduction in speed did not prevent the collision. The Mondisy, of ferro-cement construction, was of materials without inherent buoyancy and relatively unresilient. Upon impact with the CanadaSenator, the hull was breached and the sailboat quickly sank. Pleasure craft operators may not be sensitized to the fact that long voyages will require a crew that is well rested and alert and that a structured relief system is necessary. The lack of a formal post-accident pilot relief policy on the part of the Laurentian Pilotage Authority and/or the Corporation des pilotes du Saint Laurent central increases risks. On the CanadaSenator, lifebuoys were not deployed and the rescue boat was not readied and launched in an appropriate time to be effective. It is difficult to inflate a liferaft while swimming because the person in the water has to haul the painter and remain afloat. Since the liferaft could not be inflated, the crew was deprived of a primary mode of rescue.Findings as to Risk Pleasure craft operators may not be sensitized to the fact that long voyages will require a crew that is well rested and alert and that a structured relief system is necessary. The lack of a formal post-accident pilot relief policy on the part of the Laurentian Pilotage Authority and/or the Corporation des pilotes du Saint Laurent central increases risks. On the CanadaSenator, lifebuoys were not deployed and the rescue boat was not readied and launched in an appropriate time to be effective. It is difficult to inflate a liferaft while swimming because the person in the water has to haul the painter and remain afloat. Since the liferaft could not be inflated, the crew was deprived of a primary mode of rescue. An operator lacking experience and training in navigation is likely to be overreliant on the electronic chart system, a potentially risky situation when operating in restricted waterways at night or in poor visibility. Even though the sector of obscurity forward of the bow as seen from the bridge of the CanadaSenator was greater than that required by the International Convention for the Safety of Life at Sea (SOLAS), the Mondisy was seen from a distance and monitored by the bridge team; therefore the issue of visibility is not a factor in the collision. The very high frequency radiotelephone was not used to advantage by either vessel.Other Findings An operator lacking experience and training in navigation is likely to be overreliant on the electronic chart system, a potentially risky situation when operating in restricted waterways at night or in poor visibility. Even though the sector of obscurity forward of the bow as seen from the bridge of the CanadaSenator was greater than that required by the International Convention for the Safety of Life at Sea (SOLAS), the Mondisy was seen from a distance and monitored by the bridge team; therefore the issue of visibility is not a factor in the collision. The very high frequency radiotelephone was not used to advantage by either vessel. Safety Action Safety Concern Pilot Relief Subsequent to an Occurrence One of the elements of system safety in the marine environment is a pilot's performance and his/her ability to retain full concentration at all times. In order to ensure such a level of concentration, the pilot needs to be well rested and ideally emotionally removed from the occurrence. While there are provisions made to contact pilots in emergency situations and offer relief, the decision to request relief or assistance rests with the occurrence pilot. To date, there is no mandatory pilot relief policy in place at the Laurentian Pilotage Authority (LPA) nor at the Corporation des pilotes du Saint-Laurent central. However, it is reported that the LPA intends to take measures to have the pilots relieved following serious occurrences. Both fatigue and stress associated with having been involved in an occurrence may have an impact on a pilot's ability to perform his/her duties. It has long been known that stress can induce certain types of error. Finally, human nature and professional pride can hinder any objective self-assessment of a pilot's need to be relieved or assisted. In 1997, the bulk carrier Venus grounded near Bcancour, Quebec. The pilot, under contract to the LPA, had elected neither to seek relief nor to request an additional pilot to share the workload, and had remained on board for an extended period. In1999, a similar situation arose on board the grounded bulk carrier Alcor. Both the Venus and the Alcor ran aground for a second time after refloating. While not necessarily causal, degradation in pilot performance due to fatigue has been identified as a factor in the second groundings of these vessels. The TSB identified these safety issues in Marine Safety Information Letter05/01 addressed to the LPA in2001. While the need for relieving a pilot involved in an occurrence has been recognized by the Great Lakes Pilotage Authority, the LPA does not require such relief, even though it does recognize the need to relieve a pilot under normal conditions when a voyage is extended due to a slow ship or in winter conditions. In the absence of clear criteria regarding relief of pilots subsequent to an occurrence, a pilot is placed in the difficult position of making a decision on whether to request relief or assistance. Under the circumstances, a pilot may not be best suited to make this decision - a decision that can have an impact on navigational safety. The Board remains concerned that, subsequent to a serious marine occurrence, the LPA continues to operate without the benefit of a mandatory pilot relief policy, thus compromising safety and increasing risks.